Guerilla Tactics for Building Scalable E-Commerce Services with Apache Cassandra, Apache Pulsar, and Vector Search
summary
A working e-commerce prototype uses Java Spring Boot, Spring Data Cassandra on AstraDB, Apache Pulsar for order events, and a Python loader with Hugging Face + LangChain to generate 384-dim embeddings stored in a Cassandra 5.0 vector column. Product recommendations fire on add to cart and return an approximate nearest neighbor query limited to eight results, then filter out same product group. Carts use a TTL 60 days to prevent scraping bots. The live demo successfully returned relevant alternate-size and similar-item suggestions.
AI-written summary · about
chapters (14)
- 0:00Scalable e-commerce with Cassandra
- 4:13Data modeling in 2023
- 5:19Cassandra scalability features
- 6:00Spring Data Cassandra pros and pitfalls
- 7:27Pulsar for event streaming
- 8:10Vector search in Cassandra
- 9:53E-commerce platform subsystems
- 13:27Cassandra tables for shopping carts
- 15:58Product recommendation vectors
- 17:45LangChain for embeddings
- 20:35Vector search with Spring Data
- 21:59Order processing with Pulsar
- 23:44Live e-commerce demo
- 26:13Resources and learning
related talks
46mCassandra vs In Memory Data Grid in eCommerce
29mUnleashing the Power of Vector Search and Semantic Caching in Java
01h 01mVector Similarity Search in Spring
27mMachine Learning for Advanced AI Search
57mEnter The Brave New World of GenAI with Vector Search
28mA Deep Dive into Python-Powered Precision and Scalability
18mKeynote: Data = AI: Leading the Future of Gen AI with Apache Cassandra
01h 01mVector Similarity Search in Spring with Redis Stack